4.3 KiB
| title | authors | pull_requests | ||||||
|---|---|---|---|---|---|---|---|---|
| Event Split |
|
|
In past releases, all event types were defined by simply deriving the Event trait:
#[derive(Event)]
struct Speak {
message: String,
}
You could then use the various event handling tools in Bevy to send and listen to the event. The common options include:
- Use
triggerto trigger the event and react to it with a globalObserver - Use
trigger_targetsto trigger the event with specific entity target(s) and react to it with an entityObserveror globalObserver - Use
EventWriter::writeto write the event to an event buffer andEventReader::readto read it at a later time
The first two are observer APIs, while the third is a fully separate "buffered" API for pull-based event handling. All three patterns are fundamentally different in both the interface and usage. Despite the same event type being used everywhere, APIs are typically built to support only one of them.
This has led to a lot of confusion and frustration for users. Common footguns include:
- Using a "buffered event" with an observer, or an observer event with
EventReader, leaving the user wondering why the event is not being detected. On(formerlyTrigger) has atargetgetter which would cause confusion for events only meant to be used withtriggerwhere it returnsEntity::PLACEHOLDER.
Bevy 0.17 aims to solve this ambiguity by splitting the different kinds of events into multiple traits:
Event: A supertrait for observer events.BroadcastEvent: An observer event without an entity target.EntityEvent: An observer event that targets specific entities and can propagate the event from one entity to another across relationships.
BufferedEvent: An event used withEventReaderandEventWriterfor pull-based event handling.
Using Events
Events without an entity target can be defined, by deriving the BroadcastEvent trait.
#[derive(BroadcastEvent)]
struct Speak {
message: String,
}
You can then trigger the event, and use a global observer for reacting to it.
app.add_observer(|trigger: On<Speak>| {
println!("{}", trigger.message);
});
// ...
commands.trigger(Speak {
message: "Hello!".to_string(),
});
To make an event target entities and even be propagated further, you can instead derive EntityEvent.
It supports optionally specifying some options for propagation using the entity_event attribute:
// When the `Damage` event is triggered on an entity, bubble the event up to ancestors.
#[derive(EntityEvent)]
#[entity_event(traversal = &'static ChildOf, auto_propagate)]
struct Damage {
amount: f32,
}
EntityEvents can be used with targeted observer APIs such as trigger_targets and observe:
// Spawn an enemy entity.
let enemy = commands.spawn((Enemy, Health(100.0))).id();
// Spawn some armor as a child of the enemy entity.
// When the armor takes damage, it will bubble the event up to the enemy,
// which can then handle the event with its own observer.
let armor_piece = commands
.spawn((ArmorPiece, Health(25.0), ChildOf(enemy)))
.observe(|trigger: On<Damage>, mut query: Query<&mut Health>| {
// Note: `On::target` only exists because this is an `EntityEvent`.
let mut health = query.get(trigger.target()).unwrap();
health.0 -= trigger.amount();
})
.id();
// Trigger the `Damage` event on the armor piece.
commands.trigger_targets(Damage { amount: 10.0 }, armor_piece);
To allow an event to be used with the buffered API, you can instead derive BufferedEvent:
#[derive(BufferedEvent)]
struct Message(String);
The event can then be used with EventReader/EventWriter:
fn write_hello(mut writer: EventWriter<Message>) {
writer.write(Message("I hope these examples are alright".to_string()));
}
fn read_messages(mut reader: EventReader<Message>) {
// Process all buffered events of type `Message`.
for Message(message) in reader.read() {
println!("{message}");
}
}
In summary:
- Need an event you can trigger and observe? Derive
BroadcastEvent! - Need the observer event to be targeted at an entity? Derive
EntityEvent! - Need the event to be buffered and support the
EventReader/EventWriterAPI? DeriveBufferedEvent!